An experiment is a procedure that can be repeated and whose observed output is uncertain.

Let EE be an experiment. An outcome of EE is an observed output of E.E.

Let EE be an experiment. The sample space of E,E, written Ω,\Omega, is the set of all possible outcomes of E.E.

Let Ω\Omega be a sample space. An event is a subset A⊆ΩA\subseteq\Omega of outcomes. The event AA occurs if and only if the outcome ω∈Ω\omega\in\Omega of the experiment satisfies ω∈A.\omega\in A.

Let Ω\Omega be a sample space. The null event is the empty set ∅,\varnothing, which never occurs. The sample space Ω\Omega is itself an event, which always occurs.

Let Ω\Omega be a sample space. An elementary event of Ω\Omega is a singleton subset {ω}\{\omega\} of Ω\Omega that contains exactly one outcome.

Let Ω\Omega be a sample space and A⊆ΩA\subseteq\Omega be an event. The complement of AA is the event that AA does not occur.

A‾:=Ω∖A\overline{A} := \Omega \setminus A

Let Ω\Omega be a sample space and A,B⊆ΩA,B\subseteq\Omega be events. The events AA and BB are disjoint events if and only if A∩B=∅.A \cap B = \varnothing.

Let Ω\Omega be a sample space and E1,E2,…E_1, E_2, \ldots be a collection of events. The collection is pairwise disjoint if and only if Ei∩Ej=∅E_i \cap E_j = \varnothing for all i≠j.i \neq j.

Let Ω\Omega be a sample space. The set of all subsets of Ω\Omega is the power set P(Ω).\mathcal{P}(\Omega). Every event is a subset of Ω,\Omega, but not every subset of Ω\Omega is necessarily an event. The set of events F\mathcal{F} is the set of all subsets of Ω\Omega that are events, so that F⊆P(Ω).\mathcal{F}\subseteq\mathcal{P}(\Omega).

Let Ω\Omega be a sample space and F⊆P(Ω).\mathcal{F}\subseteq\mathcal{P}(\Omega). The set F\mathcal{F} is an algebra of sets on Ω\Omega if and only if it satisfies the following properties:

  • the null event is in F,\mathcal{F}, ∅∈F;\varnothing\in\mathcal{F};
  • if A∈FA\in\mathcal{F} then A‾∈F;\overline{A}\in\mathcal{F};
  • if A,B∈FA,B\in\mathcal{F} then A∪B∈F.A\cup B\in\mathcal{F}.

An algebra is closed under finite unions and finite intersections.

Let Ω\Omega be a sample space and F⊆P(Ω).\mathcal{F}\subseteq\mathcal{P}(\Omega). The set F\mathcal{F} is a σ\sigma-algebra on Ω\Omega if and only if it satisfies the following properties:

  • the null event is in F,\mathcal{F}, ∅∈F;\varnothing\in\mathcal{F};
  • if A∈FA\in\mathcal{F} then A‾∈F;\overline{A}\in\mathcal{F};
  • if E1,E2,…∈FE_1,E_2,\ldots\in\mathcal{F} then
⋃i=1∞Ei∈F.\bigcup_{i=1}^\infty E_i\in\mathcal{F}.

A σ\sigma-algebra is an algebra that is also closed under countable unions and countable intersections.

Let Ω\Omega be a set and F\mathcal{F} be a σ\sigma-algebra on Ω.\Omega. The elements of F\mathcal{F} are the measurable subsets of Ω.\Omega.

Let Ω\Omega be a set and F\mathcal{F} be a σ\sigma-algebra on Ω.\Omega. The ordered pair (Ω,F)\left(\Omega,\mathcal{F}\right) is a measurable space.

Every experiment has an associated measurable space (Ω,F)\left(\Omega,\mathcal{F}\right) where Ω\Omega is its sample space and F\mathcal{F} is a σ\sigma-algebra on Ω\Omega that contains every event of the experiment.